NeuroImage 55 (2011) 1393–1400
Contents lists available at ScienceDirect
NeuroImage
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y n i m g
Neural correlates of reality monitoring during adolescence
AnnaLaura Lagioia a,c, Stephan Eliez a,d, Maude Schneider a, Jon S. Simons e,
Martial Van der Linden b, Martin Debbané a,b,⁎
a
Office Médico-Pédagogique Research Unit, Department of Psychiatry, University of Geneva School of Medicine, Switzerland
Adolescence Clinical Psychology Research Unit, Department of Psychology and Educational Sciences, University of Geneva, Switzerland
Lemanic Neuroscience Doctoral School, Department of Psychology and Educational Sciences, University of Geneva, Switzerland
d
Department of Genetic Medicine and Development, University of Geneva School of Medicine, Switzerland
e
Department of Experimental Psychology, University of Cambridge, UK
b
c
a r t i c l e
i n f o
Article history:
Received 17 December 2010
Accepted 22 December 2010
Available online 30 December 2010
Keywords:
fMRI
Memory
Source monitoring
Schizotypy
Psychosis
a b s t r a c t
Background: Reality monitoring processes serve the critical function of discriminating between externally
derived information and self-generated information. Several reality monitoring studies with healthy adult
participants have identified the anterior prefrontal cortex (PFC) as consistently engaged during the
recollection of self-generated contextual cues. Furthermore, reduced activity of medial PFC has been linked
with schizotypal trait expression of delusion and hallucination-like symptoms in healthy adults undergoing
fMRI reality-monitoring tasks. The present study seeks to examine the cerebral underpinnings of reality
monitoring during adolescence, a developmental stage where the expression of schizotypal traits may
increase risk for psychosis.
Method: A group of 33 adolescents, assessed using the Schizotypal Personality Scale (SPQ), underwent fMRI
while performing a reality monitoring task. After an encoding session where the subject or the experimenter
read out a series of complete or incomplete word pairs, subjects were presented with the first word of studied
word pairs and asked whether the corresponding word had been: (1) perceived or produced (context
monitoring), or (2) read by the subject or by the experimenter (origin monitoring).
Results: Analyses revealed a common set of activated brain areas during both context and origin monitoring
conditions. When compared to context monitoring, origin monitoring was associated with greater activation
in anterior PFC within Brodmann area 10 (BA 10). Correlation analyses revealed that reduced signal change in
BA 10 during origin monitoring was associated with greater schizotypal trait expression.
Conclusion: Much like adults performing a similar reality monitoring task, adolescents exhibit a common
pattern of brain activity during origin and context monitoring, with functional specialization within the
prefrontal cortex involving preferential activation of BA 10 during origin monitoring. Greater schizotypal trait
expression appears to be significantly associated to reduced BA 10 activity during origin monitoring trials.
Results are discussed in relation to cortical specialization within the PFC and trait expression during
adolescence.
© 2011 Elsevier Inc. All rights reserved.
Introduction
Reality monitoring processes allow us to discriminate between
perceptually based and reflectively based mental events (Johnson and
Raye, 1981). Employing the term reality here commonly alludes to an
event's external, perceptually grounded source. When distinguishing a
past perception from a past reflection, we effectively consider the
available contextual information (modalities, features, etc.). This
process naturally contrasts with considering information of reflective
nature, such as thoughts and associated cognitive processes that more
⁎ Corresponding author. Adolescent Clinical Psychology Research Unit, Department
of Psychology and Educational Sciences, University of Geneva, 40 Boulevard du Pont
d'Arve, Case Postale 50, 1211 Geneva 8, Switzerland. Fax: +41 22 379 93 59.
E-mail address: [email protected] (M. Debbané).
1053-8119/$ – see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.neuroimage.2010.12.058
readily indicate an event's internal source (Johnson et al., 1993).
Discriminating between perceptually and reflectively based information
may be facilitated by self-relevant cues (Foley et al., 1983). In other
words, one may more promptly identify an event's perceptual or
reflective source when information on the event's origin (who delivered
the information) is available to the individual (Johnson and Raye, 1981).
Together, perceptual/reflective monitoring (hereby described as context
monitoring) and self/other monitoring (hereby described as origin
information) thus constitute two essential components of efficient
reality monitoring.
Current neuroimaging literature supports the notion that different
reality monitoring processes share neural underpinnings (Simons et
al., 2006a). Evidence for this is brought by functional neuroimaging
studies consistently reporting overlapping brain activity during origin
and context monitoring tasks, in cerebral areas that include the lateral
1394
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
prefrontal cortices (PFC), the anterior cingulate, the insula, and the
lateral parietal cortices (Simons et al., 2008). However these same
studies also highlight subtle yet critical differences in neural
activation between the two monitoring processes, showing that
origin monitoring engages significantly more activity in the rostral
prefrontal cortex within Brodmann area 10 (BA 10) compared to
cerebral activity sustained during context monitoring (Simons et al.,
2006a, 2005, 2008).
A recent conceptualization of BA 10 activity modulation argues for its
implication in the coordination of stimulus-dependent and stimulusindependent mental activity (Burgess et al., 2007). Specifically, this part
of the anterior prefrontal cortex is thought to act as a gateway regulating
the processing of perceptually or reflectively based information. Several
studies lend support to the gateway hypothesis (Gilbert et al., 2005;
Simons et al., 2006b), and more specifically, to the implication of BA 10
in reality monitoring tasks that require the participant to distinguish
between externally or internally derived information (Gilbert et al.,
2010, 2006). Moreover, because stimulus-independent thought most
likely originates from one's self, we may expect that BA 10 play a critical
role in disentangling mental content originating from the self from that
originating from another agent.
Interestingly, Simons et al. (2008) recently observed that schizotypal
personality trait expression, which is characterized by proneness to
experience delusion and hallucination-like phenomena, is associated
with reduced BA 10 activity in healthy participants performing an origin
monitoring task. Both context and origin monitoring are particularly
relevant to schizotypal trait expression, as they usually involve a
confusion between the perceptual and reflective nature of mental
events (such as command hallucinations), a misattribution of self and
other origin (such as experiencing alien thought control), or a
combination of both (such as hearing voices outside one's head). The
data reported by Simons et al. (2008) may suggest that reduced BA 10
activity during origin monitoring could, in part, reveal a faulty gateway
function increasing the propensity to experience schizotypal cognitions.
Because schizotypal trait expression constitutes the single most
predictive factor for the development of schizophreniform disorders
during adulthood (Miller et al., 2002; Poulton et al., 2000), examining
the neural underpinnings of reality monitoring during adolescence
could provide critical information on the developmental pathways
leading to minor or clinically relevant manifestations of schizotypy
(Bentall et al., 2007). Several theorists already argue for an association
between reality monitoring deficits and the expression of psychosis
(Bentall et al., 1991; Frith, 1992), and numerous reports observe
reality monitoring deficits in schizophrenic patients (Keefe et al.,
2002; Vinogradov et al., 1997), healthy adults exhibiting schizotypal
manifestations (Larøi et al., 2005, 2004), but also in diverse groups of
adolescents who report increased schizotypal trait expression
(Debbané et al., 2009a,b). To date however, the neural underpinnings
of reality monitoring processes during adolescence remain unknown.
This question entails further relevance in light of the important
cortical maturation taking place during adolescence (Casey et al.,
2008), most notably in the medial frontal cortices (Shaw et al., 2008).
Considering that psychotic psychopathology most commonly declares
itself during the early years of adulthood, it appears crucial to examine
the cognitive and neural underpinnings that may mediate schizotypal
trait expression early on during adolescence.
The present fMRI study examines context and origin monitoring in
a sample of adolescents representative of the wide range of
schizotypal trait expression. The objectives are threefold: 1) To
examine the neural underpinnings of reality monitoring during
adolescence; 2) To examine the potential overlapping and segregated
neural activations between context and origin monitoring during
adolescence; and 3) To characterize the association between
schizotypal trait expression and reality monitoring processes during
adolescence. On the basis of previous studies with adult participants,
we first expect to find overlapping cortical activation for context and
origin monitoring in our sample. Second, we hypothesize that medial
anterior PFC within BA 10 will yield significantly more activation
during origin monitoring trials when compared to context monitoring
trials. Finally, we predict that schizotypal trait expression scores will
be associated with BA 10 activation during origin monitoring trials.
Materials and methods
Participants
Thirty-three right handed teenagers (18 females) with a mean age
of 16.61 years (s.d. = 1.9), with normal or corrected to normal vision
volunteered for participation. Participants were recruited by word of
mouth from secondary schools in the state of Geneva (n = 17), and
also through a participant pool collected between 2006 and 2008 in
the child and adolescent outpatient public service, the Office MédicoPédagogique (n = 16) (Debbané et al., 2009a). At time of testing, none
of the participants were receiving any pharmacological treatment.
Two adolescents were following psychotherapeutic treatment (one
male adolescent for attentional difficulties, one female adolescent for
behavioural, attentional and academic difficulties). The Youth SelfReport and Adult Behavior Checklist (Achenbach and Rescorla, 2003)
were employed to ensure that recruitment pools were comparable and
showed no significant differences in terms of internal and external
symptoms.
All participants scored within the average range of the Wechsler
Intelligence Scale Cubes subtest. Schizotypy trait expression was
measured using the Schizotypal Personality Questionnaire (SPQ;
(Raine, 1991); French translation (Dumas et al., 2000)). This selfreport questionnaire was filled out under the supervision of a trained
clinical psychologist (M.D.). It can be applied to multiple dimensional
analyses in the context of a dimensional approach to schizotypy (Rossi
and Daneluzzo, 2002), and its full version is appropriate for use with
francophone adolescents (Badoud et al., in press). The sample's mean
total SPQ score was 18.9 (s.d. = 14.5).
Written informed consent was obtained from participants and
their parents under protocols approved by the Institutional Review
Board of the Department of Psychiatry of the University of Geneva
Medical School.
Design and procedures
Our fMRI block design procedure is adapted from Simons et al.
(2008) to evaluate context and origin monitoring in adolescents. The
task includes two phases: a study phase (encoding stage) and a test
phase (retrieval phase). The target stimuli presented during the study
phase were 96 French word pairs, following a 2 × 2 design manipulating
context (perceived versus produced) and origin (subject versus
experimenter). The encoding phase was characterized by the presentation of either both words of the pair (perceived condition), or by the
presentation only of the first word of the pair followed by a question
mark (see Fig. 1). Both the subject and the examiner were instructed to
intentionally produce and say aloud the first free associative word
coming to mind (produced condition). Both trials and conditions were
presented in a pseudo-randomised order.
The MRI sequence started during each of the three test blocks
(retrieval phase) that followed each study phase. Within the test
blocks, each subject was asked to remember the condition under
which the word pairs had been encoded during the study phase
(context and origin monitoring). Specifically they were asked to
remember whether the second word of the pair was perceived or
produced (context condition), or whether the trial was performed by
the self or the experimenter (origin condition). A third condition was
also included to provide a baseline contrast to context and origin
judgments; in this condition, new words were presented along with a
question prompting subjects to make a semantic judgment by
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
1395
fMRI data analysis
Fig. 1. Reality monitoring task adapted from Simons et al. (2008).
classifying words as either naturally occurring or man-made objects.
Participants were asked to answer by pressing one of the two buttons
of an MRI compatible button box.
The encoding and retrieval phases were divided into 3 separate
sessions to optimize memory load. Each encoding phase was
constituted by 32 target trials immediately followed by the retrieval
phase. An encoding session required either the subject or the
examiner to read aloud the word pair appearing on the screen, with
trials organized in a pseudo-randomised fashion, so that the same
individual could undertake no more than 3 consecutive trials.
Each retrieval session immediately followed the encoding session
and involved 12 blocks equally divided into the 3 conditions:
condition A (words perceived or produced), condition B (words
performed by the self or the experimenter), and condition C (baseline
condition natural/man made objects). Two rest blocks (R – continuous
fixation cross) started and closed the test sessions. Each retrieval
phase consisted of 48 target trials (32 word pair in addition 16
naturally occurring or man made objects).
Block sequence for the first two phases was R-ACABBCAACBCB-R
whereas for the last retrieval phase conditions AB were inverted. Each
block was made up of 4 target words. A 4000 ms instruction display
preceded trials, preceded by a fixation cross of 500 ms. The target
stimuli were presented and answer given within 4500 ms, with an ITI
of 100 ms was separating trials. Each block lasted 24 s, totalling 6 min
and 28 s for each test phase, for a total time of approximately 28 min
dedicated to the task.
Data were processed and analyzed using Statistical Parametric
Mapping 5, (SPM5; Welcome Department of Neuroscience, London,
UK). Functional images were first corrected for motion by realigning
all images with respect to the first image acquired; re-sampling of all
slices was run to match the middle slice of the entire acquisition to
correct for differences in slice acquisition timing; each participant's
structural image was coregistered to the mean of the realigned
functional images. Grey matter separation was established by
segmentation of the anatomical image. Realigned and slice timed
images were then spatially normalised into the Montreal Neurological
Institute template (MNI) using a 3-mm cubic voxel size and finally
spatially smoothed with a 6-mm isotropic Gaussian Kernel full width
half maximum (FWHM). A high pass filter of 1/128 Hz was used to
remove low frequency noise, and an AR (1) + white noise model
corrected for temporal autocorrelation. For each participant brain
responses were estimated at each voxel using a general linear model
with block regressors. The 24 sec block lengths relative to the period
of appearance of the words of the context memory and baseline
assignment were convolved with the canonical hemodynamic
response function in the first statistical analysis. The design matrix
included the realignment parameters to account for any enduring
movement.
Experimental contrasts were performed to obtain subject specific
estimates for each of the effects of interest using the general linear
model as implemented in SPM 5 resulting in a voxel value map of t
statistics for 12 contrasts. The threshold for subject images was set at
an uncorrected p b 0.05. These images were further smoothed (6-mm
Gaussian Kernel) to conform to inter-individual brain size variability
for one-sample t test group analysis.
Statistical parametric maps for each group contrast were characterized
by an uncorrected height threshold of p b 0.001. The peak locations of
local maxima cluster of contiguous voxels were localised on a mean
structural scan with approximate Brodmann areas estimated from the
Talairach and Tournoux (1988) atlas after having adjusted coordinates
to allow for differences between the MNI and Talairach templates
(www.mrc-cbu.cam.ac.uk/Imaging/Common/mnispace.shtml).
To enhance exploration differences between the controls and
patients' group a region of interest (ROI) analysis was run using
Marsbar (http://marsbar.sourceforge.net/). The ROI was obtained
using an unbiased contrast-based whole-brain analysis, to insure that
potential correlations between this ROI and the independently
obtained schizotypy measures would not represent an artifact of
ROI selection. A 6 mm radius sphere was designed around the centre
of mass for each subject to extract statistical tables. A Pearson
correlation was run to explore the relationship between brain activity
and behavioural patterns particularly between context memory of the
origin and source of the stimuli and schizotypy measures.
Results
Behavioural results
fMRI data acquisition
Blood oxygenation level-dependent (BOLD) functional images
[parameters: interscan repetition time (TR) = 2400 ms, echo time
(TE) = 30 ms, slice thickness = 3.20 mm, flip angle = 85°, FOV
235 mm] and high resolution three-dimensional anatomical images
[TR = 2500 ms, TE= 30 ms, slice thickness = 1.1 mm, flip angle= 8°,
192 volumes, FOV 220 mm] were acquired using a 3 T Siemens TIM Trio
system. Three scan sessions each consisting of one hundred and sixtyfour volumes were acquired while subject performed the word pair task.
Each volume comprised 38 slices oriented parallel to the AC-PC line and
collected in a descending sequence.
Participants' performance for context and origin trials are reported
in Table 1. Comparisons related to condition performances revealed
significantly superior retrieval performances in the origin condition
compared to the context condition (t(32) = −6.12, p b 0.001). Within
the origin condition, we observed a significant difference in favour of
experimenter trials (t(32) = 6.54, p b 0.001) in comparison to self trials.
In the context condition, we observed significant difference in favour of
performances on seen trials (t(32) = 3.70, p = 0.001) in comparison to
produced trials. To investigate possible Encoding × Retrieval interactions, we performed two repeated-measure ANOVAs with post-hoc
Tukey pairwise comparisons. This yielded that encoding context
significantly influenced origin retrieval (F(3,96) = 44.92, p b 0.001;
1396
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
Table 1
Total performance accuracy on the word-pair reality monitoring task.
Conditions
Accuracy — percentage
correct (s.d.)
Origin trials
(retrieval-encoded conditions)
Self trials
Self-seen
Self-produced
Experimenter trials
Experimenter-seen
Experimenter-produced
Context
(retrieval-encoded conditions)
Seen trials
Seen-self
Seen-experimenter
Produced trials
Produced-self
Produced-experimenter
Semantic baseline
77.02 (9.16)
Reaction times —
milliseconds (s.d.)
1971.2 (306.08)
70.20
57.07
82.82
83.83
79.29
88.63
59.85
(11.79)
(19.22)
(9.97)
(10.00)
(12.34)
(10.16)
(14.40)
1938.0 (305.22)
1957.8 (327.45)
1867.9 (471.38)
2002.72 (343.86)
2168.81 (386.3)
1833.49 (367.04)
2279.83 (302.65)
69.95
71.8
66.91
49.75
48.98
50.75
94.25
(16.49)
(14.7)
(23.6)
(25.17)
(28.16)
(25.54)
(6.09)
2296.56 (332.57)
2121.40 (498.55)
2401.4 (402.9)
2261.54 (345.36)
2291.48 (443.25)
2237.5 (354.35)
1535.3 (297.49)
Table 2b
Cerebral regions showing greater activation during retrieval of Self/Experimenter
stimuli compared to semantic baseline condition. Coordinates are in MNI atlas space
(Cocosco et al., 1997), and brain regions estimated from the Talairach and Tournoux
(1988) atlas after normalization (Talairach and Tournoux, 1988).
Anatomical label
Hemi
BA
MNI (x, y, z)
t (value)
Superior parietal lobule
Cuneus
Inferior parietal lobule
Precuneus
Middle frontal gyrus
R
L
L
R
R
L
L
L
L
R
R
R
R
R
L
R
R
L
L
R
L
R
7
7
40
7
9
46
6
6
8
10
11
10
9
8
8
8
36, −63, 54
−6, −69, 36
−36, −51, 42
6, −69, 39
45, 27, 33
−42, 21, 27
−36, 6, 57
−35, 3, 39
−6, 18, 48
36, 60, 3
27, 45, −9
21, 66, −3
39, 12, 42
36, 15, 54
−6, 18, 48
6, 24, 48
33, 24, −6
−54, −33, 0
−53, −33, −9
69, −33, −9
−21, −87, −18
27, −84, −3
12.99
10.23
9.95
9.42
9.98
8.02
9.11
7.94
7.51
4.94
4.48
3.53
8.09
5.71
7.51
6.29
7.11
5.83
4.92
5.32
5.51
5.60
Precentral gyrus
Superior frontal gyrus
Medial frontal gyrus
Insula
Middle temporal gyrus
Experimenter × Context: p = 0.009; Self × Context: p = 0.0001). However, encoding origin did not significantly influence context retrieval
(F(3,96) = 10.38, p b 0.001; Seen × Origin: p = 0.65; Produced × Origin:
p = 0.98).
Origin and context condition comparisons for reaction time
revealed only one significantly faster retrieval performance, suggesting
faster origin retrieval compared to the context retrieval (t(32) =-7.09,
p b 0.001). We again performed two repeated-measure ANOVAs with
post-hoc Tukey pairwise comparisons to investigate possible
Encoding×Retrieval interactions regarding reaction times. This yielded
that encoding context significantly influenced origin reaction time
but only when the origin was external (F(3,96)=12.57, pb 0.001;
Experimenter×Context: p=0.0001; Self×Context: p=0.44). Furthermore, encoding origin significantly influenced context reaction time but
only for seen encoding trials (F(3,96)=11.93, pb 0.001; Seen×Origin:
p=0.0001; Produced×Origin: p=0.81).
Neuroimaging results
Reality monitoring conditions versus baseline contrasts
Brain regions associated with the two types of reality monitoring
task were firstly analyzed by contrasting activation maps of each
condition against the semantic baseline (Table 2a). Large areas of
bilateral anterior prefrontal cortex (PFC), dorsolateral PFC, middle
Table 2a
Cerebral regions showing greater activation during retrieval of Seeen/Produced stimuli
than semantic baseline condition. Coordinates are in MNI atlas space (Cocosco et al.,
1997), and brain regions estimated from the Talairach and Tournoux (1988) atlas after
normalization.
Anatomical label
Hemi
BA
MNI (x, y, z)
t (value)
Middle frontal gyrus
L
L
L
R
L
L
L
L
R
R
R
R
L
R
R
6
46
9
9
10
21
22
8
6
32
7
40
7
−36, 3, 57
−45, 18, 27
−51, 24, 30
48, 27, 36
−36, 54, 6
57, −39, −9
−54, −42, 0
0, 21, 48
3,27,39
12, 36, 21
36, −66, 54
45, −54, 54
−6, −66, 36
33, 24, −6
39, 6, 54
12.39
10.05
9.46
9.69
7.70
6.98
6.04
9.61
9.23
5.85
12.36
11.40
10.95
10.79
9.77
Middle frontal gyrus
Middle temporal gyrus
Medial frontal gyrus
Anterior cingulated
Superior parietal lobule
Inferior parietal lobule
Precuneus
Insula
Superior frontal gyrus
8
Fusiform gyrus
Middle occipital gyrus
22
21
21
18
18
temporal gyrus (MTG) and lateral parietal cortices were activated
commonly during both conditions. Some differential areas of
activation were also observed. When contrasting the Origin (Self/
Experimenter) condition against the Semantic baseline, the peak
activation [p b 0.001 (unc.)] was registered bilaterally around the
superior parietal lobe, cuneus followed by the engagement of the right
middle frontal gyrus and medial frontal regions. When contrasting the
activations maps for Context condition (Seen/Produced) against the
semantic baseline, activity [p b 0.001 (unc.)] peaked around the right
middle frontal gyrus and parietal areas and precuneus (Tables 2b, 2c,
and 2d; Fig. 2).
Comparing origin and context monitoring and their links to schizotypal
trait expression
When contrasting the two reality monitoring conditions that
represent origin and context monitoring, significant differences were
observed around the postcentral gyrus and in medial frontal gyrus in
the Self/Experimenter N Seen/Produced contrast (Fig. 3), showing
increased activity in the left medial PFC (BA10 centred on −6, 69, 3;
t = 3.58, p b 0.001). When Seen/Produced condition was contrasted
against Self/Experimenter differences were underlined around the
middle temporal gyrus and superior temporal regions.
To explore the interaction between the medial PFC activation and
schizotypal trait expression, we performed a post-hoc correlation
analysis between the participants' total SPQ score and signal change
Table 2c
Cerebral regions showing greater activation during retrieval of Seen/Produced
compared to Self/Experimenter condition. Coordinates are in MNI atlas space (Cocosco
et al., 1997), and brain regions estimated from the Talairach and Tournoux (1988) atlas
after normalization.
Anatomical label
Hemi
Pyramis
Middle temporal gyrus
L
L
R
R
R
L
R
Superior frontal gyrus
Cingulate
Medial frontal gyrus
BA
MNI (x, y, z)
t (value)
21
46
9
8
32
9
−9, −87, −33
69, −33, −1
54, 33, 27
51, 21, 30
36, 21, 21
−12, 21, 33
6, 36, 33
4.31
4.20
3.91
3.64
4.02
3.37
4.01
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
Table 2d
Cerebral regions showing greater activation during retrieval of Self/Experimenter
condition compared to Seen/Produced condition. Coordinates are in MNI atlas space
(Cocosco et al., 1997), and brain regions estimated from the Talairach and Tournoux
(1988) atlas after normalization.
Anatomical label
Hemi
BA
MNI (x, y, z)
t (value)
Postcentral gyrus
Medial frontal gyrus
R
R
L
L
L
R
L
L
L
3
6
6
6
10
13
13
24
11
15, −42, 72
9, −30, 69
−12, −33, 63
−12, −24, 54
−6, 69, 3
27, −30, 18
−36, 15, 15
−9, 0, 48
−33, 45, −9
4.61
4.14
3.72
3.58
3.61
4.22
3.48
3.81
3.58
Insula
Cingulate gyrus
Middle frontal gyrus
around the peak activation observed in the Self/Experimenter N Seen/
Produced contrast. A 6 mm radius sphere was implemented around the
coordinates of cluster of interest (x= -6, y = 69, z = 3). Signal change
during recollection of origin information was extracted on a subject by
subject basis. The correlation analysis between signal change in the
cluster of interest and total SPQ score revealed a significant a correlation
(r = 0.329, p b 0.001) (Fig. 5). Significant correlations were found with
the discrete dimensions evaluated by the SPQ, positive symptoms sub
scale (r = 0.283, p = 0.001), negative symptoms subscale (p = 0.001,
r = 0.308) and disorganized dimension (r = 0.263, p = 0.002).
To ensure that medial PFC activation was not a product of task
performance (accuracy and reaction time) or age, we ran Pearson
correlations between extracted medial PFC signal change values and
both accuracy percentage and reaction times on origin trials across
participants. Results did not show any significant associations with
accuracy scores (r = 0.002, p = 0.798), reaction time (r = 0.009,
p = 0.605), or age at participation (r = 0.05, p = 0.199). The analyses
reveal no correlation between the ROI signal change and the other
two conditions (Context condition r = −0.231, p = 0.246; Control
condition r = −0.220, p = 0.270). The results were not due to
variability at an individual level; calculating signal change error
scores for each participant (% of signal change for origin N context) and
corresponding individual behavioural error score (% of origin
accuracy N context accuracy) lead to a non-significant correlation
between these two scores (r = −0.002, p = 0.827).
Finally, in order to ensure that our main findings concerning the
specificity of BA 10 activation for origin trials and its significant
relationship to adolescent schizotypal trait expression were not
driven by poor behavioural performances in the context condition,
we eliminated the subjects with a context (n = 7) or origin (n = 0)
total accuracy score below 50%. Performing our contrast of interest on
1397
the reduced sample (n = 26) showed that our cluster of interest in BA
10 remained significant (BA10 centred on −6, 69, 3; t = 2.61,
p b 0.005 unc.). Furthermore, the correlation between signal change
and total SPQ score remained strongly significant (r = 0.302,
p = 0.004). These additional analyses provide important support to
the reliability of the present results (Fig. 4).
Discussion
Results from the fMRI reality monitoring paradigm administered to
a sample of adolescents provide support to the hypotheses tested in
this study. First, we observed that different reality monitoring
processes, namely origin and context monitoring, largely share
underlying neural activation patterns including the anterior prefrontal
cortex (PFC), the dorsolateral PFC, the MTG, the posterior cingulate
gyrus and both medial and lateral regions in the parietal cortex.
Second, we observed that adolescents exhibit significantly more
activity in left medial BA 10 during retrieval of self/experimenter
information compared to retrieval of contextual information. Finally,
we found that schizotypal trait expression during adolescence is
negatively associated with left BA 10 engagement during self/
experimenter trials. We will first briefly discuss the behavioural
results and how they may inform us about reality monitoring skills
during adolescence. Second, we will focus on the implications of these
results for understanding the functional specialization of the PFC
during adolescence with regard to previous studies performed with
adults. Finally, we shall discuss the specific results relating left BA 10
activation to the expression of schizotypal traits during the adolescent
developmental period.
Performance on our reality monitoring paradigm suggested that
adolescents find retrieving the origin of an event (self or experimenter) to be an easier task compared to retrieving its contextual
characteristics (was the event seen or produced). This observation
is consistent with Foley et al.'s (1983) seminal study on children's
memory for speech and thought. In that study, the authors conducted
a series of experiments highlighting that the ability to discriminate
between one's own speech and another person's speech is developmentally mature as early as age 6, while the ability to distinguish
between what one said and what one only thought undergoes further
elaboration during development. Therefore, different developmental
courses may underlie the monitoring of contextual cues (seen versus
produced), which seem to yield a maturational advantage for
distinguishing origin information (self versus not self), as observed
in our study. Consistent with this view, young adults past the stage of
adolescence (aged 19–36 years) undergoing the same fMRI paradigm
showed less of a discrepancy (0.87 versus 0.8) between the two
Fig. 2. Activation regions during retrieval of Origin N Semantic Baseline stimuli (left); Activation regions during retrieval of Context N Semantic Baseline stimuli (right).
1398
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
Fig. 3. Activation regions during retrieval of Origin N Context stimuli (left); Activation regions during retrieval of Context N Origin stimuli (right).
reality monitoring conditions (Simons et al., 2008). Additional studies
employing a longitudinal design will help identify the progressive
reinforcement of context monitoring abilities along with their
potential relationships with the cortical maturation that characterizes
the adolescent developmental period (Shaw et al., 2008). As an aside,
it is worth noting that the differential task performance exhibited by
our adolescent sample is unlikely to be a sufficient explanation for the
neuroimaging results obtained, as no significant correlations were
observed between anterior PFC signal and behavioural measures.
When examining the cerebral activity sustaining reality monitoring in our adolescent sample, we find strikingly similar activity
patterns as those found in adult participants performing the same task
(Simons et al., 2008). As with adults, we can witness a pattern of
overlapping cerebral activity between origin and context monitoring.
Moreover, contrasting origin with context monitoring reveals a
comparable functional specialization within the anterior prefrontal
cortex in adolescents, bringing confirmatory evidence that increased
BA 10 activity is specifically related to origin monitoring. These
observations suggest that the cerebral networks sustaining reality
monitoring processes are already in place during adolescence. The
lack of correlation between local specializations of BA 10 with age
may appear surprising, given the important maturational processes
targeting the prefrontal cortex during adolescence (Shaw et al., 2008).
One possible explanation could be that while the functional
components of reality monitoring already resemble those observed
in adulthood, their specialization and efficiency may still be
Fig. 4. Medial view of signal change in left BA 10 for the Origin N Context contrast (x =
−6 y = 69 z = 3).
maximised through further maturation during adolescence and into
adulthood, as suggested by additional regions of cerebral activation in
adolescent origin monitoring, as well as poor adolescent behavioural
context monitoring performances. Future analyses examining functional connectivity linking the anterior prefrontal cortex to other
cortical areas may help uncover the more refined maturation of reality
monitoring processes through adolescent development.
With respect to our final hypothesis, we observe a significant
relationship linking reduced BA 10 signal change to the adolescents'
reported schizotypal trait expression. The results can be interpreted
from two main theoretical perspectives, namely the source monitoring framework (SMF; (Johnson et al., 1993)) and the gateway
hypothesis (Burgess et al., 2007). The SMF suggests that one performs
source judgments of self-generated material based on the available
information associated to the self-relevant mental events. This
information may primarily consist in perceptually based information
(contextual details associated to the event) or reflectively based
information (cognitive operations performed on mental content); the
first type of information will more likely lead to evaluating the mental
event as coming from an external source, while the latter will more
likely be perceived as originating from an internal source. In a recent
update of the SMF integrating the latest neuroimaging findings
relevant to source monitoring, specific activation of BA 10 is thought
to be involved in the evaluation of internally-generated information,
as well as the active retrieval and selection of pertinent information
(Mitchell and Johnson, 2009). Our interpretation may further be
informed by our behavioural results, which suggest that produced
information increases the likeliness of retrieval success during selfexperimenter trials. We may point out that wordpairs produced by
adolescents contain both internally (private cognitive operations
linked to the production of the wordpair) and externally based
information (one's externally voiced response), whereas self-seen
wordpairs involve the voiced response while lacking the richness of
cognitive operations linked to wordpair production. Given the
increased performances for self-produced in comparison to self-seen
trials, we may conjecture that evaluating internal self-generated
information during origin trial appears critical to successful performances. This would be consistent with the SMF account of BA 10
activity as specifically sensitive to the evaluation of internallygenerated information. In turn we may suggest that adolescent failing
to evaluate internally self-generated information would also tend to
express more important levels of schizotypy. This interpretation
would be consistent with the accumulated evidence showing
impaired recalling self-generated information in psychosis-prone
and schizophrenic adults (Brunelin et al., 2007; Larøi et al., 2005).
Concurrently, the gateway hypothesis may lead us to consider that
a failure to evaluate internally-generated information might be
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
1399
especially to François Lazeyras, Pascal Challande and Frank Henry. We
would also like to acknowledge Michal Epstein and Deborah Badoud
for their contributions to data collection. This work was funded by the
Gertrude Von Meissner Foundation (ME 7871) grant to S. Eliez and M.
Debbané, and by the Swiss National Fund (PP00B-102864) grant to S.
Eliez.
References
Fig. 5. Correlation between signal change extracted from the cluster of interest within
the medial PFC (BA 10) during retrieval of self/experimenter trials and total SPQ score
(r = 0.329, p b 0.001).
further supported by inefficient modulation of attending to perceptually
versus reflectively based information. Activation of more anterior part of
the medial prefrontal cortex has been associated with attending to
perceptual details (Gilbert et al., 2007) and memory for task (Simons
et al., 2005), nearby the BA 10 coordinates that we found discriminated
the origin from the context monitoring conditions in our adolescent
sample (x= −6, y = 69, z = 3). This suggests that the modulation of
attention to externally versus internally generated information during
the origin monitoring trials also recruits BA 10 activity during
adolescence. Furthermore, the expression of schizotypal traits would
also be promoted by inefficient regulation of the gateway managing
attention between externally and internally derived information. This
view is consistent with several accounts describing psychotic-like
symptoms as self-generated mental events attributed to an external
source (Bentall et al., 1991; Frith and Done, 1989). Characteristically,
auditory hallucinations constitute internally generated mental events
whose origin is experienced as greatly questionable, and whose
reflective basis can be misinterpreted as perceptual. More nuanced
occurrences infiltrate reflective mechanisms such as alien thought
control, which again entails a confusion between mental content
originating from within but experienced as under the control of an
external agent. Interestingly, the cognitive processes described above
may promote the expression of schizotypal trait expression through
adolescent development. Future studies should examine the developmental antecedents of both reality monitoring and the attentional
gateway to gain further insight into the unfolding of psychotic-like
phenomena.
Together, these results point to the relevance of reality monitoring
mechanisms during adolescence, and their possible implication in
personality trait expression such as schizotypy. They further lend
cross-sectional developmental evidence to contemporary cognitive
accounts implicating self-monitoring deficits in psychotic-like experiences (Frith, 1992). They further suggest that inconsistencies in
origin monitoring may constitute discrete cognitive alterations that
contribute to individual differences in schizotypal trait expression.
Origin monitoring processes could be further examined in high affect
situations, which are known to be conducive to maintenance and
exacerbation of early schizotypal cognitions (Escher, 2004; Escher
et al., 2002). Finally, future investigations may examine whether
origin monitoring processes could represent a primary prevention
target in youths at high-risk for the unfolding of psychotic disorders.
Acknowledgments
The authors would like to thank the volunteer participants, as well
as Drs Dario Balanzin and Serges Djapo-Yogwa for their collaboration.
Additional thanks go to the CIBM/LAVIM neuroimaging platform,
Achenbach, T.M., Rescorla, L.A., 2003. Manual for the aseba adult forms and profiles.
University of Vermont, Research Center for Children, Youth, and Families,
Burlington.
Badoud D., Chanal J., Eliez S., Van Der Linden M., and Debbané M. (in press). Validation
study of the french schizotypal personality questionnaire in an sample of
adolescents; a confirmatory factor analysis. L'Encéphale.
Bentall, R.P., Baker, G.A., Havers, S., 1991. Reality monitoring and psychotic hallucinations. Br. J. Clin. Psychol. 30 (Pt 3), 213–222.
Bentall, R.P., Fernyhough, C., Morrison, A.P., Lewis, S., Corcoran, R., 2007. Prospects for a
cognitive-developmental account of psychotic experiences. Br. J. Clin. Psychol. 46
(Pt 2), 155–173.
Brunelin, J., d'Amato, T., Brun, P., Bediou, B., Kallel, L., Senn, M., et al., 2007. Impaired
verbal source monitoring in schizophrenia: an intermediate trait vulnerability
marker? Schizophr. Res. 89 (1–3), 287–292.
Burgess, P.W., Dumontheil, I., Gilbert, S.J., 2007. The gateway hypothesis of rostral
prefrontal cortex (area 10) function. Trends Cogn. Sci. 11 (7), 290–298.
Casey, B.J., Jones, R.M., Hare, T.A., 2008. The adolescent brain. Ann. NY Acad. Sci. 1124,
111–126.
Cocosco, C.A., Kollokian, V., Kwan, R.K.S., Evans, A.C., 1997. Brainweb: online interface to
a 3D mri simulated brain database. Neuroimage 5, 425.
Debbané, M., Van der Linden, M., Gex-Fabry, M., Eliez, S., 2009a. Cognitive and
emotional associations to positive schizotypy during adolescence. J. Child Psychol.
Psychiatry 50 (3), 326–334.
Debbané, M., Van der Linden, M., Glaser, B., Eliez, S., 2009b. Monitoring of self-generated
speech in adolescents with 22q11.2 deletion syndrome. Br. J. Clin. Psychol. 49 (3),
373–386.
Dumas, P., Bouafia, S., Gutknecht, C., Saoud, M., Dalery, J., d'Amato, T., 2000. validation of
the french version of the raine schizotypal personality disorder questionnaire—
categorial and dimensional approach to schizotypal personality traits in a normal
student population. Encephale 26 (5), 23–29.
Escher, S., 2004. Determinants of outcome in the pathways through care for children
hearing voices. Int. J. Soc. Welf. 13 (208–222).
Escher, S., Romme, M., Buiks, A., Delespaul, P., Van Os, J., 2002. Independent course of
childhood auditory hallucinations: a sequential 3-year follow-up study. Br. J.
Psychiatry Suppl. 43, s10–s18.
Foley, M.A., Johnson, M.K., Raye, C.L., 1983. Age-related changes in confusion between
memories for thoughts and memories for speech. Child Dev. 54 (1), 51–60.
Frith, C.D., 1992. The Cognitive Neuropsychology of Schizophrenia. Psychology Press
Ltd., Hove.
Frith, C.D., Done, D.J., 1989. Experiences of alien control in schizophrenia reflect a
disorder in the central monitoring of action. Psychol. Med. 19 (2), 359–363.
Gilbert, S.J., Frith, C.D., Burgess, P.W., 2005. Involvement of rostral prefrontal cortex in
selection between stimulus-oriented and stimulus-independent thought. Eur. J.
Neurosci. 21 (5), 1423–1431.
Gilbert, S.J., Spengler, S., Simons, J.S., Steele, J.D., Lawrie, S.M., Frith, C.D., et al., 2006.
Functional specialization within rostral prefrontal cortex (area 10): a metaanalysis. J. Cogn. Neurosci. 18 (6), 932–948.
Gilbert, S.J., Williamson, I.D., Dumontheil, I., Simons, J.S., Frith, C.D., Burgess, P.W., 2007.
Distinct regions of medial rostral prefrontal cortex supporting social and nonsocial
functions. Soc. Cogn. Affect. Neurosci. 2 (3), 217–226.
Gilbert, S.J., Henson, R.N., Simons, J.S., 2010. The scale of functional specialization within
human prefrontal cortex. J. Neurosci. 30 (4), 1233–1237.
Johnson, M.K., Raye, C.L., 1981. Reality monitoring. Psychol. Rev. 88, 67–85.
Johnson, M.K., Hashtroudi, S., Lindsay, D.S., 1993. Source monitoring. Psychol. Bull. 114
(1), 3–28.
Keefe, R.S., Arnold, M.C., Bayen, U.J., McEvoy, J.P., Wilson, W.H., 2002. Sourcemonitoring deficits for self-generated stimuli in schizophrenia: multinomial
modeling of data from three sources. Schizophr. Res. 57 (1), 51–67.
Larøi, F., Van der Linden, M., Marczewski, P., 2004. The effects of emotional salience,
cognitive effort and meta-cognitive beliefs on a reality monitoring task in
hallucination-prone subjects. Br. J. Clin. Psychol. 43 (Pt 3), 221–233.
Larøi, F., Collignon, O., Van der Linden, M., 2005. Source monitoring for actions in
hallucination proneness. Cogn. Neuropsychiatry 10 (2), 105–123.
Miller, P., Byrne, M., Hodges, A., Lawrie, S.M., Owens, D.G., Johnstone, E.C., 2002.
Schizotypal components in people at high risk of developing schizophrenia:
Early findings from the edinburgh high-risk study. Br. J. Psychiatry 180,
179–184.
Mitchell, K.J., Johnson, M.K., 2009. Source monitoring 15 years later: what have we
learned from fmri about the neural mechanisms of source memory? Psychol. Bull.
135 (4), 638–677.
Poulton, R., Caspi, A., Moffitt, T.E., Cannon, M., Murray, R., Harrington, H., 2000.
Children's self-reported psychotic symptoms and adult schizophreniform disorder:
a 15-year longitudinal study. Arch. Gen. Psychiatry 57 (11), 1053–1058.
Raine, A., 1991. The SPQ: a scale for the assessment of schizotypal personality based on
DSM-III-R criteria. Schizophr. Bull. 17 (4), 555–564.
1400
A. Lagioia et al. / NeuroImage 55 (2011) 1393–1400
Rossi, A., Daneluzzo, E., 2002. Schizotypal dimensions in normals and schizophrenic
patients: a comparison with other clinical samples. Schizophr. Res. 54 (1–2), 67–75.
Shaw, P., Kabani, N.J., Lerch, J.P., Eckstrand, K., Lenroot, R., Gogtay, N., et al., 2008.
Neurodevelopmental trajectories of the human cerebral cortex. J. Neurosci. 28 (14),
3586–3594.
Simons, J.S., Gilbert, S.J., Owen, A.M., Fletcher, P.C., Burgess, P.W., 2005. Distinct roles for
lateral and medial anterior prefrontal cortex in contextual recollection. J.
Neurophysiol. 94 (1), 813–820.
Simons, J.S., Davis, S.W., Gilbert, S.J., Frith, C.D., Burgess, P.W., 2006a. Discriminating
imagined from perceived information engages brain areas implicated in schizophrenia. Neuroimage 32 (2), 696–703.
Simons, J.S., Scholvinck, M.L., Gilbert, S.J., Frith, C.D., Burgess, P.W., 2006b. Differential
components of prospective memory? Evidence from fmri. Neuropsychologia 44
(8), 1388–1397.
Simons, J.S., Henson, R.N., Gilbert, S.J., Fletcher, P.C., 2008. Separable forms of reality
monitoring supported by anterior prefrontal cortex. J. Cogn. Neurosci. 20 (3),
447–457.
Talairach, J., Tournoux, P., 1988. Co-Planar Stereotaxic Atlas of the Human Brain. Georg
Thieme Verlag, Stuttgart, New York.
Vinogradov, S., Willis-Shore, J., Poole, J.H., Marten, E., Ober, B.A., Shenaut, G.K., 1997.
Clinical and neurocognitive aspects of source monitoring errors in schizophrenia.
Am. J. Psychiatry 154 (11), 1530–1537.